R Tutorial: Nonlinear Modeling in R with GAMs | Intro

Want to learn more? Take the full course at https://learn.datacamp.com/courses/nonlinear-modeling-in-r-with-gams at your own pace. More than a video, you'll learn hands-on coding & quickly apply skills to your daily work. --- Hi, I'm Noam Ross. I'm a scientist who studies infectious diseases. I use R and Generalized Additive Models to better understand complex biological and ecological systems. This course will teach you how to use these versatile models to analyze and understand complex, multifaceted, non-linear relationships in your own work. Whenever we build statistical models, we face a trade-off between flexibility and interpretability. GAMs offer a middle ground between simple models, such as those we fit with linear regression, and more complex machine learning models like neural networks. Linear models are easy to interpret and to use for inference: It is easy to understand the meaning of their parameters. However, we often need to model more complex phenomena than can be represented by linear relationships. On the other hand, machine learning models, like boosted regression trees or neural networks, can be very good at making predictions of complex relationships. The problem is that they tend to need lots of data, are quite difficult to interpret, and one can rarely make inferences from the model results. GAMs offer a middle ground: they can be fit to complex, nonlinear relationships and make good predictions in these cases, but we are still able to do inferential statistics and understand and explain the underlying structure of our models and why they make predictions that they do. GAMs let us flexibly model non-linear relationships. Here I've made a scatter plot of two variables, X and Y. We can see from the scatterplot that there is clearly some relationship between the variables, but it is not linear. If we fit a linear model to the data using the lm() function and the usual formula syntax, we can see it won't do a very good job. The